Projects

Open-source projects in AI, software engineering, web tools, and digital transformation by Ricardo Squassina Lee.

A selection of original projects, practical tools, and technical contributions. Projects are grouped by theme; each entry links to its source code and, when available, its live application.

AI and intelligent systems

Agent Flow

An environment to register LLM-based agents, chain them into flows, and track executions and token usage. The repository includes a full version built with Docker, FastAPI, Streamlit, and SQLite, as well as a static Lite version that runs in the browser.

auto-model

A GitHub event router that selects AI models based on task requirements and the balance between capability and cost. It covers code reviews, documentation, issue triage, quick comments, and release notes, with GitHub Actions integration.

PromptGen

A web tool that turns plain-language answers into structured prompts for writing, planning, research, analysis, and coding. It includes guided examples, a quality indicator, and an editable generated prompt.

CV-JD-Match

A client-side tool that compares a resume with a job description using LLMs, provides a structured match analysis, and can generate a tailored resume and cover letter. Communication with the AI provider happens directly from the browser.

Web products and tools

TechJobScout-lite

A client-side tool for technology professionals to explore opportunities, assess job fit, and prepare STAR interview stories, with no backend or database.

XadrezMestre

An open-source chess training application built with React and TypeScript, using the Stockfish engine to play against the computer.

Fator R Simulator

A web simulator for comparing Brazilian Simples Nacional Annex III and V scenarios, considering the Fator R ratio, owner compensation, DAS, INSS, and IRRF. Calculations run in the browser.

**Note:** Fator R Simulator is an educational and planning tool. Results are estimates, tax parameters must be checked against current legislation, and decisions should be validated with a qualified tax professional.

Engineering, delivery, and collaboration

TPM Toolbox

A proposed project-delivery process combining Shape Up, Microsoft’s Code With Engineering Playbook, and Agile practices. It explores planning, shaping, prioritization, technical spikes, build cycles, and reviews. The repository describes the material as an evolving proposal, not a finalized methodology.

Documentation CI Pipeline

A quality pipeline for technical documentation, combining Markdown standards, link checking, and writing clarity analysis. It demonstrates automation of engineering and documentation review practices.

AI-in-a-Box

A collaborative collection of AI and machine learning accelerators and guidance, including Responsible AI, GenAI application security, MLOps, and Azure deployment patterns. This is a collaborative repository derived from a Microsoft Customer Engineers and Architects community initiative; it is not presented as an exclusively individual-authored project.


Back to home